Department of Comparative Politics

New Article

Do Polls Influence Opinions?

A new publication by Sveinung Arnesen, Mikael P. Johannesson, Jonas Linde and Stefan Dahlberg examines the potential self-fulfilling prophecy of public opinion.


Linde, Dahlberg,Arnesen and Johannesson.
The four researchers have been working on their newly published article for a long time, and are excited to be able to shed some light on an issue debated for decades.

Through a novel automated procedure labeled the dynamic response feedback (DRF), the four researchers measured whether the answers from early poll respondents can influence the opinions of subsequent respondents who learn the answers of the previous respondents. Whether respondents were affected by poll influence, was tested on their attitudes towards Syrian refugees and mandatory measles vaccination, after having learned the result of previous opinion polls on the same issues. The experiments were implemented in waves four and five of the Norwegian Citizen Panel, a probability-based research-purpose online survey panel administered by the Digital Social Science Core Facility at the University of Bergen.

The work leading up to the publishing of this article, has been going on for a long time. Arnesen, one of the authors, explains that the issue of polls, and the effect of polls such as the bandwagon effect, is always relevant. In some countries, polls are not allowed to be published in the final two weeks before elections. Through the Norwegian Citizen Panel, there was a technical opening to research this phenomenon, which has been of interest for both social scientists and the general public for a long time.

Worst Case Scenario: The Bandwagon Effect

Poll effects—and in particular the bandwagon effect—have been debated for a long time, a mechanism where exposure to polls affect public opinion in the direction of the majority. This risk was first referred to in electoral studies. It has since been applied beyond the electoral context, but there are relatively few studies on the influence of polls in nonelection settings, and they demonstrate contradictory results. In addition to the bandwagon effect, there is the spiral of silence hypothesis, which claims that citizens who perceive their attitudes to be in the minority refrain from expressing their views publicly out of fear of social isolation; consequently, the dominant majority will therefore become more dominant and louder over time. This would be counterproductive to the ideals of democratic deliberation, where public discourse between free and equal individuals take place, not peer pressure. 

Positive Negative Findings

The results from the four researchers’ DRF experiment show that when the respondents are exposed to a poll that shows other people’s opinions on the issue, the aggregate level impact on public opinion is negligible. There is therefore little reason to worry that polls are disruptive to public debate and deliberative processes. 

While public opinion is, overall, little affected by individuals’ exposure to information about the majority opinion, they nevertheless observe that some of those who see the polls on vaccination adjust their perceptions of others’ opinions, which, in turn, changes their own attitudes on the issue. And, they change their opinions in the same direction as their perceptions of public opinion, which is an indication of a bandwagon effect.

- Bandwagon effects have been studied for decades, says Arnesen. Our innovation is that we have created an experimental design which efficiently captures the time dimension of such effects in surveys. The results from our experiments presented in the paper show that any influence of polls on other polls do not sustain over time. 

In the future, the researchers would like to pursue similar investigations, where they use this experimental procedure to test whether their initial findings hold also for oher issues and contexts.

- For example, we want to use the procedure to test in-group and out-group influence, that is, whether people are more likely to be influenced by the poll if the poll dispalys the attitudes of people the respondents identify with. There is reason to believe that people listen more closely to the opinions of people they identify with.

It has traditionally been easier to publish positive findings, which can often appear more sensational. Arnesen says it was a bit more challenging to get an article with a null result published. But for the sake of research credibility, it’s important not to dramatize findings.

- It’s a null result, but it’s an interesting null result, contributing to our knowledge by weakening a common belief, says Arnesen.